strategy-dev
Paper trade → live capital: a 30-day promotion plan
The structured ramp from virtual P&L to real money — without overconfidence
30 days (5 min/day check-ins) · intermediate
What you'll have when finished
- Live strategy with validated edge across multiple market conditions
- Confidence that came from data, not gut feel
- Built-in scale-up schedule that prevents over-betting on a hot streak
- Clear go/no-go criteria so you stop bad strategies fast
Before you start
- A strategy that worked in paper trading can still fail live — markets shift
- The biggest mistake is scaling up too fast after a hot first week
- Skip ANY step here and you're back to gambling
Walkthrough
-
Day 0: Backtest + Strategy Coach pass (1 hour)
Before paper trading, validate the basics: 1. **Backtest** your strategy on the last 90 days. Look for Sharpe >1, max drawdown <25%, profit factor >1.3, win rate consistent with your sample. 2. **Strategy Coach review.** Fix every critical issue. Decide on warnings (some are fine, some matter). 3. **Document your hypothesis.** Write down: "I expect this strategy to make X% per month, with Y max drawdown, on Z market conditions." This is your falsifiable claim. If backtest fails, don't paper trade. Iterate on the strategy until backtest passes.
Success criteria: Backtest Sharpe > 1 · Profit factor > 1.3 · Strategy Coach: zero critical issues · Written hypothesis with target P&L and drawdown
-
Days 1-7: Paper trade with REAL planned size
Open /paper-trading. Configure: - **Starting capital = your planned live deployment** (e.g., $10K virtual if you plan to deploy $10K real) - **Fee model matches your exchange** (Coinbase ~0.4%, Binance ~0.1%) - **Slippage 0.1%** on majors, 0.3% on alts Run for 7 days minimum. Check in once per day for 2-3 minutes. DO NOT panic-close paper trades. End-of-week 1 check: Did the strategy fire at least 10 trades? If fewer, the sample is too small — run another week.
Success criteria: Paper session size matches planned live size · 7+ days elapsed · 10+ trades in the session · No manual interference (let it run)
-
Days 7-14: Compare paper vs backtest, iterate if needed
Side-by-side compare: - Paper win rate vs backtest win rate — within 10%? - Paper Sharpe vs backtest Sharpe — within 30%? - Paper max drawdown vs backtest — paper shouldn't exceed backtest by more than 50% **Big divergence (>30% off) = something is different live.** Common causes: - Slippage you didn't model - API latency causing late fills - A regime change since backtest period If divergence is small, continue paper trading another week to confirm. If divergence is large, go back to backtest and figure out why.
Success criteria: Paper vs backtest metrics compared · Divergence < 30% on key metrics · Or root cause identified and addressed
-
Day 14: Promote to LIVE at 25% size
Open the real bot config. Set position size to **25% of your planned full size**. Why 25%? Because: 1. Live conditions still differ from paper in subtle ways 2. 2 weeks of paper trading is not "proven" — you've validated the math, not the longevity 3. Starting small means a discovered bug costs you 25% as much Ensure: - Risk Rules has the auto-halt set - Daily loss limit accounts for the smaller live size (proportional) - You'll journal every trade for the next 30 days
Success criteria: Live position size = 25% of plan · Risk Rules auto-halt enabled · Journal entry for every trade
-
Days 14-21: First 7 live days — match paper or pull back
Compare LIVE results to PAPER results week-over-week: ✅ **Live ≈ Paper:** Strategy holds up. Continue at 25% size. ⚠ **Live worse but not catastrophic:** Hold at 25%, another week of data. ❌ **Live much worse:** Pause. Find the cause. Don't scale up. Hot streak warning: if first 5 trades are big winners, DO NOT increase size yet. Variance up = variance down later. Stay disciplined to the schedule.
Success criteria: Daily check-ins (5 min) · Journal entry per trade · Stay at 25% even on a hot streak
-
Day 21: Scale to 50% IF live matches paper
Only proceed if your live P&L over the first 7 live days is within 30% of your paper-trade results. If yes: increase position size to 50% of planned full size. Continue daily check-ins. If no: stay at 25% another week. Some strategies need 14+ live days for variance to smooth out. Common mistake: scaling up because the dollar P&L is good, even though the strategy underperformed paper. Don't reward bad relative performance just because absolute numbers are positive.
Success criteria: Live vs paper within 30% · Live position size = 50%
-
Day 30: Full size IF the strategy proved itself
By day 30 you have 16 days of live results at 25-50% size. If all of the following are true, scale to 100%: - Sharpe is positive (any Sharpe < 0.5 means stop) - Max drawdown is within your Risk Rules limit - You've had at least one losing day and didn't override the strategy - Live profit factor > 1.2 on 20+ trades If any of these fail, STAY at the current size for another 30 days. The cost of patience is some opportunity. The cost of impatience is the account. This whole process saves you from being the trader who blew up a great strategy by scaling 10× after one good week.
Success criteria: Sharpe > 0.5 in live · Max drawdown within limits · 20+ live trades · Discipline maintained through at least one losing day
What's next
Three things separate traders who survive year 2 from those who don't: 1. **Quarterly stress test.** Run [Stress Test](/portfolio-analytics?tab=stress-test) against your current allocation every quarter. Strategies that worked in 2024 may not work in 2025 conditions. 2. **Re-backtest after regime shifts.** When BTC dominance changes, when stables de-peg, when major news hits — your strategy's edge may have changed. Re-validate. 3. **Critical journal review.** Once a month, read the last 30 trades. Are you still following the strategy? Or has discretionary creep set in? Discretionary creep kills systematic strategies. A live strategy is a living thing. The 30-day promotion plan gets it deployed; ongoing discipline keeps it alive. **Next:** [Build a risk-managed system](/learning-library?guide=build-risk-managed-trading-system) if you haven't already.